Chapter 2
Error and Tolerance
Abstract This chapter addresses the conceptual relations between error and tolerance in the context of adjudicating, through crucial moments in the formulation of
rabbinic legal thought. It outlines the conceptual approaches to the tolerable-intolerable boundary rooted in Biblical categories, defined differently by the Talmudic
sages and further transformed by medieval jurists. While the Talmudic literature
offers three articulations of the distinction between tolerable and intolerable judicial
error—substantive, conventional and scholastic—post-Talmudic jurists confront
the very notion of judicial error and reflect critically on the Talmudic definitions.
Since the concept of error is connoted with epistemology and that of tolerance with
political philosophy, the notion of tolerable error in a jurisprudential context inhabits a territory in which two disciplines intersect, to ask under what circumstances
judicial errors may be tolerated and on what grounds tolerating certain errors may
be justifiable.
The concept of tolerance in modern times is labyrinthine and not easily encapsulated, despite being supposedly necessary to a multi-cultural existence; the same
is true when discussing tolerance in pre-modern religious contexts. The study of
tolerance may be at its most effective by following not a bivalent structure as much
as a sequence, from the affirmable to the tolerable to the intolerable. Generally
speaking, these three qualities represent the three possible predicates attributed to
all kinds of deeds and beliefs under any theory of tolerance. At one extreme, what is
affirmable includes all acts or beliefs which are totally consistent with the set of beliefs of the individual who is tolerant, and hence are fully legitimized by that belief
set. At the other extreme, intolerable acts or beliefs are wrongs which one cannot
accept and live with. Between the affirmable and the intolerable lie acts and beliefs
with which one is prepared to live despite objections.
Two boundaries in this sequence are of great importance to the concept of tolerance and its operation. The first is the boundary between what is affirmable and
what is tolerable. To understand this boundary, we must develop a clear sense of
what constitutes the distinction between affirmable and tolerable: If each reflects a
degree of acceptance, to what extent is this distinction tangible? Will what is tolerable ultimately become affirmable? The second is the tolerable-intolerable boundary, which invites exploration of what constitutes the difference between various
opposed wrongs: what is it that allows one to tolerate one wrong but reject another?
J. E. David, Jurisprudence and Theology, Studies in the History of Law and Justice 2,
DOI 10.1007/978-3-319-06584-7_2, © Springer International Publishing Switzerland 2014
11
12
2 Error and Tolerance
Is what differentiates them conventional, contextual, or essential, or does it depend
on other perspectives?1
The following discussion will sketch the historical outlines of the concept of
tolerable error within Jewish legal thought and examine this concept against parallel notions in contemporary Islamic jurisprudence which may have affected Jewish
thinkers.2 This sketch will follow a chronological order, tracing the development
of notions of tolerable error as they were perceived by the rabbis. By accounting
for the two boundaries of the tolerable discussed above, this discussion aims to illuminate some features of the idea of tolerance in Jewish intellectual history in this
period.
The concept of tolerable error in Jewish legal tradition from the Mishnah to
the Middle Ages can be traced back to the biblical distinction between inadvertent
sin ( ‫ )שגגה‬and deliberate sin ( ‫)זדון‬. The measure of wrongdoing in this context is
therefore dependent on intention, as in the standard assessment of criminal acts: in
criminal law, the objective details of an act ( actus reus) are necessary but not sufficient to judge criminality, and as long as the act is not accompanied by an intention
to act wrongly ( mens rea), some Tolerance of the act is possible. In biblical law, the
standard of intention is further expanded to various types of misconduct.3 Thus, one
may view the origins of the notion of tolerable error as an expansion of the concept
of inadvertence to include cognitive failure as well as behavior.4 An expression of
this expansion may be seen in the following homily of a tannaitic midrash, which
expands the category of inadvertence to include cognitive fallacies:5
And if (they) commit an inadvertent error.6 might one suppose that they are liable solely
on account of the inadvertent commission of deed? (Nay) The Scripture says: that has
erred and the matter escapers the notice of the congregation—They are liable only on
Some writers emphasize that the concept of tolerance in fact rests on skepticism, rationalism or
pluralism, see McKinnon (2006). The connection between tolerance and skepticism is also evident
in pre-modern and non-western traditions, see Shagrir (1997).
2 The distinction between tolerable and intolerable error is rooted in the Roman legal tradition as
well. The maxim error juris nocet, error facit non nocet (derived from the Dig. 22.6 De iuris et
facti ignorantia) led to the distinction between ‘error of law’ and ‘error of fact’ which was traditionally perceived to reflect the distinction between inexcusable and excusable errors (hence the
canonist maxim error juris non excusat). It was suggested by Keedy that the original rationale
behind the distinction made in the Justinian’s Digest is based on the Romans’ perceptions of legal
norms and their intelligibility. Accordingly, they held that ‘the law is certain and capable of being ascertained, while the construction of facts is difficult for even the most circumspect’ (Keedy
1908, p. 78). The Romans believed ‘that law can and should be fixed, whereas facts tend to be
more elusive. Owing to the greater elusiveness of facts, error as to them should be excused more
readily so that only supine ignorance or crass negligence should bar excuse’ (Ryu 1957).
3 In fact, the transition from ‘sin’ to ‘error’ is already hinted in biblical verses when describing the
inadvertent act in cognitive terms—ignorance of the matter (‘… and that thing was hidden from
the eyes of the congregation and they did’, Lev. 4:13) as opposed to knowledge of the sin (‘… and
the sin which they committed was known…or his sin was made known to him’, Lev. 4:14).
4 Evans (2002, pp. 11–19).
5 SifLev 4:13.
6 Lev. 4:13.
1 2 Error and Tolerance
13
account of something’s being hidden along with an act of transgression which is performed
inadvertently.
The entire project of the Mishnaic tractate Horayot, (‫הוריות‬, lit. decisions or rulings)
which discusses norms pertaining to judicial errors in rabbinic courts, can be seen as
an implementation of the biblical concept of inadvertent sin within rabbinic institutions, an exegetical move that applies the biblical theory of action to the rabbinic
theory of adjudication.7
In the biblical system, the tolerable-intolerable boundary was thus perceived as
a problem of agency determined by a human measure—the actor’s state of mind,
intention and mental condition—but in rabbinic law from the Talmud onwards, intention as a criterion for tolerable error was not further developed. Instead, there
was a trend to seek external means of assessing which errors can be tolerated. In
other words, Talmudic articulation of tolerable errors devalues the weight of the
wrongdoer’s intention and increases reliance on objective standards.8
The Talmud contains three different approaches to the phenomenon of tolerable
error within the context of adjudication, each expressing a different conceptualization of the tolerable-intolerable boundary. These approaches can be characterized
as follows: (1) According to the ‘substantive’ approach, tolerable error is defined
according to the relationship between the corpus juris, the core body of the law, and
its peripheral aspects. The concept of tolerable error thus reflects a fundamentalist
perspective on the law, for which reason every deviation from the fundamentals is
considered illegitimate. (2) According to the ‘conventional’ approach, the tolerable-intolerable boundary is determined by cases involving contravention of laws
regarding religious ideas or behaviour that all Jews can reasonably be expected to
share. (3) According to the ‘scholastic’ approach, tolerable error is related to the
broad rabbinic project of rearticulating the notion of divine law and identifying it
with the teachings of the sages. Of these approaches, the first finds expression in the
statements of tannaim, mainly in the laws of the Mishnaic tractate Horayot, while
the other two approaches are to be found first in amoraic literature, albeit possibly
reflecting earlier conceptions.
The substantive approach to the notion of tolerable error is structured, from the
perspective of the law, by a core of norms (in rabbinic terms: the essence of the
law—i’kar hadin—or its body—guf hadin) and its particulars. A court’s ignorance
or error with regard to essential parts of the corpus juris constitutes an excessively
serious deviation which cannot be tolerated:9
The characterization of tractate Horayot as representing an anti-Sadducee worldview is supported by the last article, which explicitly manifests an anti-priestly ideology—‘A priest takes
precedence over a Levite, a Levite over an Israelite, an Israelite over a bastard ( ‫ …)ממזר‬this order
of precedence applies only when all these were in other respects equal. If the bastard, however, was
a scholar and the high priest an ignoramus (‫)עם הארץ‬, the learned bastard takes precedence over the
ignorant high priest.’ (m Hor. 3:8).
8 If Talmudic thought was shaped in reaction to contemporary Christians, as some have argued,
one might associate this with reactions to the promotion of Christian values of inwardness.
9 m Hor. 1:3.
7 14
2 Error and Tolerance
(If the) court ruled that an entire body (of the law) has to be uprooted; if they said, (for
example), that (the prohibition against having intercourse with) menstruating woman is
not found in the Torah or the (prohibition against labour on) Sabbath is not found in the
Torah or (prohibition against) idolatry is not found in the Torah—lo, they are exempt (from
bringing sin-offering).
(If, however,) they ruled to nullify a part (of a law) and a part retained, they are liable (and
thus should bring sin-offering).
How so?—if they said: (the prohibition against having intercourse with) menstruating
woman is indeed in the Torah but he who has sexual relations with a woman awaiting day
against day is exempt, (or that prohibition against labor on) Sabbath is in the Torah but he
who carries out something from private domain to public domain is exempt, (or that the
prohibition against) idolatry occurs in the Torah, but a man who bows down (to an idol)
is exempt,—lo, these are liable; for scripture says, ‘and a matter escapes the notice (of the
assembly)’10—‘a matter’, but not the entire body.
The Mishnah, here and later on, articulates the tolerable/intolerable distinction in
terms of atonable/non-atonable errors. Any grave error that relates to the essence
of the law cannot be atoned, and therefore anyone who commits such a mistake
is exempt from bringing a sin-offering, since it would be useless to make such an
offering.11 A tolerable error thus encompasses any deviation that does not violate
the integrity of the entire law. The more the law is distorted by a court’s erroneous
ruling, the less the erroneous ruling is tolerated. If the court’s ruling indeed rejects
the essence of the law (or its entire body), the ruling is considered void and thus not
subject to tolerance at all. If, on the other hand, the ruling is not in conflict with the
essential components of the law but only with its particulars, that erroneous verdict
can be tolerated. This conception expresses a deep concern for the integrity of the
law,12 hence the tolerable-intolerable boundary is located at that point where the erroneous ruling is likely to distort the entire law. Conversely, tolerable error extends
over the realm of deviations which do not nullify the body of the law.
The conventional approach to tolerable error is expressed by reference to the
alleged degree of consensus on some religious matters among Jews, including the
notoriously ‘heretic’ Sadducees, in Second Temple times. The defining principle of
the tolerable-intolerable boundary according to this approach is therefore ‘a matter
which the Sadducees admit’—that is, fundamentally, those undisputed matters
Lev. 4:13.
The terminology of this distinction is also widespread in the Talmud’s laws relating to an idolatrous prophet. In this context it serves to emphasize the seriousness of deviation from the essential laws, which is understood as distorting the Torah itself. The metaphor of ‘uprooting a thing’
indicates the fixed and stable nature of the Torah, denial or nullification of which are seen as a
distortion of the original Torah.
12 The conception of ‘Law as Integrity’, as introduced in Ronald Dworkin’s legal theory, states
that the law must speak with one voice, so that judges must assume that the law is structured on
coherent principles and that in all fresh cases which come before them, judges must apply the same
principles. Accordingly, integrity is both a legislative and an adjudicative principle. The legislative
principle requires law makers to try to make the laws morally coherent. Law-makers are required
assume that these principles were created by, or for, the community as an entity, and that they
express the community’s conception, see Dworkin (1986, p. 225).
10 11 2 Error and Tolerance
15
about which the rabbis had no record of a disagreement between Pharisees and the
Sadducees:13
Rav Judah said in the name of Samuel: The court is not liable (for the erroneous ruling)
only when they ruled concerning a prohibition which the Sadducees do not admit; but if
concerning a prohibition which the Sadducees admit, they are exempt.
It follows from this that the tolerable-intolerable boundary does not pertain either to
the laws themselves or to the extent to which the Torah, as divine law, is distorted,
but rather to the degree of agreement with the Sadducees’ legal norms or rulings.
The precise significance of the use of the tag ‘a matter which the Sadducees
admit’ as a standard is obscure. There are at least two possible ways of reading the
phrase. If the Sadducees were viewed as holders of erroneous knowledge but still
holding to certain elementary norms which corresponded to the law acknowledged
by other Jews, these norms would be considered by the rabbis as basic and therefore
‘obligatory knowledge’, so that whoever errs regarding them could not be tolerated.
Alternatively, a ‘matter which the Sadducees admit’ might indicate norms in a territory of consent between the Sadducees and other groups, such as Pharisees, and
these norms might be considered such common knowledge that no-one could ignore
or miss them.
The standard of a matter ‘which the Sadducees admit’ also appears in Talmudic
discussion concerned with judicial errors in trials involving death sentences. Among
the differences between civic suits (in Talmudic terms, ‘monetary laws’) and suits
involving capital punishments, is a rule which grants procedural privilege to the
defendant in capital suits—‘monetary suits may be reversed either for acquittal or
conviction; capital cases may be reopened for acquittal, but not for conviction’.14
As opposed to civic cases, the defendant in capital suits enjoys the benefit of an erroneous acquittal. The Talmud, however, limits this tannaitic principle according to
the same standard:15
‘But not for conviction’. R. Hiyya b. Abba said in R. Johanan’s name: proving that he erred
in a matter which the Sadducees do not admit. But if he erred in a matter which (even) they
admit, (we tell him16) go back to school and learn it.
How does the admission of the Sadducees about a specific matter mark the tolerable-intolerable boundary? According to the Talmud, this question was already asked
by the tannaim of the school of R. Ishmael:17
R. Shesheth replied, and so it was taught by the School of R. Ishmael: Why has it been said
of a court that ruled ‘a matter which the Sadducees admit’, that they are exempt? Because
they should have learned and did not learn.
b Hor. 4a.
m Sanh. 4:1.
15 b Sanh. 33b.
16 This addition appears in: b Hor. 4a.
17 b Hor. 4b.
13 14 16
2 Error and Tolerance
Accordingly, basic knowledge, which even the Sadducees admit, is obligatory for
any jurist. Lack of such knowledge is intolerable; an unforgivable negligence that
cannot be atoned by sin-offering. The tolerable-intolerable boundary here is determined by an a priori obligation. Thus the concept of tolerable error stands in relation to the concept of negligence. The rebuke ‘Go back to school and learn it’ is of
course semantically close to the admonition ‘because they should have learned and
did not learn’. Both identify the boundary of tolerable error with negligence–that is,
ignorance of fundamental legal knowledge.
The scholastic approach to tolerable error, like the conventional approach, also
appears in the amoraic layer of the Talmud and is not found in tannaitic statements.
The Talmud suggests a distinction between tolerable error, or error regarding judicial discretion ( ta’ut beshiqqul hada’at; ‫)טעות בשקול הדעת‬, which should not be
reversed if it occurs, and error regarding an explicit teaching of the sages ( ta’ut
bedevar mishnah; ‫)טעות בדבר משנה‬, which can be reversed and considered a cause
for compensation if damage occurs:18
R. Shesheth said in R. Assi’s name: If he erred in regarding devar mishnah,
the decision is reversed; if he erred in shiqqul hada’at, the decision may not be
reversed.
The distinction drawn between these two categories—devar mishnah19 and
shiqqul hada’at—is rooted in understanding of the authority of the sages and the
canonical status granted to their teachings. In fact, this conception abolishes the gap
between jurists and legal norms. Contrary to viewing the sages as merely the authorized commentators on the law, this perception presents an idea of self-canonization
which provides the sages’ teachings with an authoritative status in their own right.20
Accordingly, the sages are not only the mouthpieces of the law or its discoverers,
but rather they constitute the law by their teachings and rulings. The casuistic definitions of these categories underline the scholastic perception by which the law is
b Sanh. 33a.
The accepted view is that the term mishnah (teaching) was introduced following the Destruction
of the Second Temple, even before it was applied to a specific corpus which came to be known
as the Mishnah of R. Judah the Prince. The literal meaning of this term is evidently related to the
activity of repetition; nevertheless, it also indicated the idea of independent authority ascribed to
the sages. Compared with the early midrashic literary style which presents the laws as an outcome
of a correct reading of the scriptures, the mishnaic style represents the laws as an inseparable part
of the sages and their scholastic activities. The independent authority of the Mishnah was taken
as equivalent to the scriptures and might be reflected in the Greek form of the term as used by
the Church fathers (δεντερωσις) in the various translations of the verb ‫ה‬.‫נ‬.‫ש‬, to repeat (δεντεροω),
or ‘the Sages taught’ (οι σοφοι δεντερωσιν), and in anonymous Geonic commentary quoted by
Nathan ben Jehiel of Rome, Aruch Completum: the reason ‘it is called Mishnah, is because it is
secondary (‫ ;שניה‬second) to the Torah … And it is the Oral Torah; and it is clear that it is secondary
to the first thing.’
20 A commentator in fact speaks for the interpreted text and therefore his authoritativeness is derivative and secondary to the authority of the text. The idea of self-canonization implies the view
according to which the sages take themselves as firsthand authorities. Therefore the difference
between the two perceptions of the sages’ authority illustrates the shifting from the text, as source
of authority, to the sage, and from interpretation to teaching.
18 19 2 Error and Tolerance
17
identified with the teachings of the sages. Hence, a deviation from the sages’ teachings by a court is intolerable and thus can be reversed and subjected to compensational remedies.21 On the other hand, when a judge deviates from those teachings
that are not explicitly fixed, but only determined by second order principles,22 this
error can be tolerated and his decision must be allowed to stand.
Of these three Talmudic approaches, only the scholastic approach continued to
develop in post-Talmudic law. It seems that, for whatever reason, the other two approaches were not deemed useful conceptions of the tolerable-intolerable boundary
in later legal thought.
A tradition attributed in a later source to the last head of the Pumbeditan academy, Hai Gaon (939–1038), states that ‘every report (‫ )שמועה‬of all the scholars is
considered devar mishnah,’23 so that knowledge was reckoned to expand beyond
the Talmudic material, embracing the entire community of rabbinic scholars.24This
view was not accepted by all. The Provençal Talmudic scholar Zerahiah b. Isaac
ha-Levi (1125–1186) recorded his disagreement with an anonymous scholar25 who
had argued for the inclusion of the teachings of the Babylonian authorities—the
Geonim—within the category of devar mishnah as part of the obligatory knowledge
in respect of which erring is intolerable:26
And I have heard, in the name of one of the scholars in the previous generation, that in our
times no one errs in shiqqul hada’at, for all our halakhic rulings are fixed, either in the Talmud or by the post-Talmud scholars. Therefore, in our times, no one errs in shiqqul hada’at,
but all those who err, err in devar mishnah. And I do not think so, but whoever makes an
error not clearly conflicting the Mishnah or the Talmud is undoubtedly not erring in devar
‘Ravina asked R. Ashi: Is this also the case if he erred regarding a teaching of R. Hiyya or R.
Oshaia? Yes, said he, And even in a dictum of Rab and Samuel? Yes, he answered. Even in a law
stated by you and me? He retorted, Are we then reed cutters in the bog? How are we to understand
the term: shiqqul hada’at?—R. Papa answered: If, for example, two Tannaim (i.e. sages of the
Mishnah) or Amoraim (i.e. sages of the Talmud) are in opposition, and it has not been explicitly
settled with whom the law rests, but he (the judge) happened to rule according to the opinion of
one of them, whilst the general practice; follows the other,—this is a case of (an error) in shiqqul
hada’at.’ (b Sanh. 33a). On the status of such second order guiding principles in unresolved disputations see the our discussion below.
22 The two examples brought in the Talmud are the spirit of the Talmudic discussion and the
widespread habit.
23 This statement is documented only in later sources and it is probably known only through the
testimony of Estori Ha-Parhi (Isaac b. Moses; 1280–1355) who learned about it through his travels
in the Orient, see Kaftor va-Ferah, Chap. 12, p. 51.
24 The idea of which the general acceptance of a legal corpus makes it obligatory and of a canonical status renounce the notion of ijma’ ( ωΎϤΟ· ) in Islamic jurisprudence. Ijma’ refers to the
consensus of the Muslim community. The various schools within Islamic jurisprudence define the
consensus differently: Some argued that it refers to the consensus of the first generation or the first
three generations of Muslims only; some held that it refers to the consensus of the jurists in the
Muslim world in general and some even extend it to include consensus of all the Muslims, scholars
and laymen alike.
25 Ta-Shma identifies this scholar with Joseph b. Meir ibn Migash (1077–1141) the head of the
yeshiva in Lucena, see Ta-Shma (1992, p. 113).
26 Sefer ha-Maor, 80b.
21 18
2 Error and Tolerance
mishnah, but rather erring in shiqqul hada’at. … and what has been ruled by the discretion
of the Geonim, subsequent to the sealing of the Talmud, and does not conflict with a clear
Talmudic norm, is like a general practice, and whoever errs regarding that, his error is of
shiqqul hada’at and not of devar mishnah.
While the older scholar opposed by Zerahiah apparently suggested viewing the category of devar mishnah as including all the rabbinic teachings up to his days, Zerahiah ha-Levi limited the category only to Talmudic teachings. In our terms, the two
scholars argued about how to sketch the tolerable-intolerable boundary with regard
to post-Talmudic rulings. The former tended to expand the range of intolerable errors, whilst the latter widened the realm of the tolerable.
One move away from the scholastic approach which took place in Geonic legal thought may be related to the development of Islamic jurisprudential ideas and
methods, since the new meaning ascribed to tolerable error by these rabbis reflected
a wholehearted change in their basic understanding of the concept of law and the act
of adjudication. The change arose in the aftermath of the emergence of reasoningbased jurisprudence, that acknowledges and praises the involvement of legal reasoning in interpreting the law and applying it to new cases.
The ideas expressed by Sherira b. Hanina27 about judicial errors were based on
the Talmudic typology of tolerable and intolerable errors, but the meaning that he
ascribed to these categories reflected a remarkable departure from Talmudic notions
and a deep absorption of Islamic legal concepts into Jewish legal thought. In this
respect, Sherira’s embrace of Islamic jurisprudential concepts completely modified
the traditional setting of the law and the meaning of legal reasoning. Consequently,
he provides innovative accounts of what the law is, what adjudication is, and what
judicial error concerns. Following the conceptual vocabulary of Islamic law, he departs from the scholastic perception of tolerable error and instead favours an objectivist approach which rests on a view of the law as made up of roots and branches.
Judicial reasoning, accordingly, is about drawing analogical links between existing
laws, ‘roots’, and new cases, ‘branches’:28
In one of these two things judges err: either this legal case has a root, (which has) a tradition
or ruling, and this judge did not know it has some resemblance (to that root), and (instead
he) analogizes it to a different root—by that he errs in devar mishnah.
Or else, (when) this case is definitely a branch that has nothing in common with (another
root), and that judge analogizes it to a root, which is not similar and has nothing in common—by that he errs in shiqqul hada’at.
This account of judicial error ignores the notion of error as a deviation from, or contradiction of, the teachings of the sages: for Sherira, judicial error is a failure to draw
correct analogical links. The difference between tolerable and intolerable errors is
therefore articulated according to a botanic metaphor and the relations of roots and
branches. By viewing judicial error as a fallacy in applying analogical reasoning,
Sherira expresses a position that seeks to impose constraints on its use and to limit
Born around 900 and died around 1000. He served as the head of the Babylonian academy
( yeshiva) in Pumbedita, which was relocated to Baghdad towards the end of the ninth century.
28 Otzar ha-Geonim, Sanh., pp. 25–26.
27 2 Error and Tolerance
19
the range of legal solutions that may be obtained thereby. Not only does he reject
the view of analogical reasoning as a product of a jurist’s personal preference, but
he also proposes a new approach according to which executing judicial analogy
requires substantive correlations between the root and its branches. Such an attitude
limits the range of possible outcomes. In fact, Sherira wishes to create a conceptual
criterion to distinguish between valid and erroneous analogies.29 Such a perception is best understood as an objectivist approach.30 We can summarize Sherira’s
position as an endeavour to provide relevant meaning to the Talmudic typology of
judicial error in accordance with the Islamic theory of qiyas.31 Such relevance is
possible with an objectivist attitude which allows the peculiar juxtaposition of ‘erroneous analogy’.
Now we turn to the other boundary of the framework, that between what can be
affirmed and what can only be tolerated. Compared with the tolerable-intolerable
distinction, it is more difficult to account for the difference between the affirmable and the tolerable, since it does not reflect a division in principle between the
legitimate and the illegitimate: in practice, both the affirmable and the tolerable are
accepted, and thus the difference between them tends to be blurred. Indeed, we can
sometimes trace the steps by which the distinction between the affirmable and the
tolerable has been blurred both in practice and in theory.
Examples of this shift were expressed independently by two dominant Jewish
jurists of the twelfth century: Isaac b. Abba Mari (1122–1193) and Moses Maimonides (1135–1204). Both scholars nullify, on different grounds, the Talmudic
category of discretional error ( ta’ut beshiqqul hada’at), so that what had previously
been considered only tolerable was deemed by them to be affirmable.
First, Isaac b. Abba Mari introduced a different perspective on the multiplicity of
opinions in the Talmud. Against the perspective of the Talmud, Isaac b. Abba Mari
explicitly ignored any second order procedure in deciding between opposing teachings of the sages. To him, multiplicity of teachings, and even opposing teachings,
should be a source of legitimization for multiple practices:32
Sherira’s insights can be compared to those of Sayf al-din al-Amidi (d. 1233). For Amidi, the
resemblance between two analogized cases is anticipated due to its preexistence and not the aftermath the jurist’s deliberation. By that, he situates the qiyas outside of the sphere of intellectual
activity of the jurists; see Weiss (1992, pp. 552–553).
30 It seems that for Sherira ‘roots’ and ‘branches’ have slightly different denotations. They do not
indicate the principles and their particular derivations, but rather stand for two types of resemblance. A ‘root’ is a legal norm whose potential similarities, subject to further analogies, are fixed
in advance. Therefore, failure to construe the ab initio similarities is an intolerable judicial error. A
‘branch’ is a legal norm whose potential resemblances are not prefixed. An analogy drawn to this
norm does not undermine preexisting resemblances and is thus tolerable.
31 The Arabic term qiyas (αΎϴϗ ) in its legal sense refers to judicial analogy, general deduction or
syllogism. At times legal qiyas is considered the archetype of all forms of legal argumentation. In
particular, it indicates the various types of argumentations that legal scholars use in their independent reasoning. The term probably originated in the ancient Hebrew term hekesh (‫)היקש‬, based on
the Aramaic root n.k.sh which means to ‘hit together’ See also p. 89 below.
32 Itturei Soferim, pp. 157–158.
29 20
2 Error and Tolerance
What are (the circumstances of) shiqqul hada’at like? If two Amoraim are mutually
opposed, and sugia deshma’ta (lit. a tradition on that matter), corresponds to the (other)
one, it is not (considered) an error for whoever follows the practice of one to do so, and
whoever follows the practice of the other to do so.
In fact, Isaac b. Abba Mari does not acknowledge any aspiration to legal unity in
the first place, since for him, the open texture of the Talmud is sufficient for full
acceptance, and there is therefore no need to have to tolerate what should be fully
affirmable.33
Isaac’s contemporary, the great thinker and jurist Moses Maimonides, also tried
to eliminate the affirmable-tolerable boundary, though on different grounds. As a
typical rationalist in theology and in law, Maimonides, like the Babylonian Geonim,
supported the use of legal reasoning. When dealing with the Talmudic typology of
judicial error, he displayed much sensitivity with regard to the wide range of possibilities achieved by embracing legal reasoning. His approach on this issue expresses
a sophisticated combination of fundamentalism34 on the one hand, and reductionism
on the other, hence his interpretation of the Talmudic categories of judicial error:35
First, I will explain that judicial error may occur in one of two things, either with reference
to (authoritative) transmitted text, as when he forgot the language or did not learn it, and
this is called an error in devar mishnah. And the second is when he errs in a thing dependent
on analogy: if the thing is possible as he stated, nevertheless the (common) practice contradicts it, and this is called an error in shiqqul hadaat.
Maimonides, like Sherira, adopts Islamic legal theory by equating the Talmudic
categories to the distinction between revealed law ( al-nass) and its extension by
means of analogy ( qiyas)36 but, unlike Sherira, at least prima facie, he remains loyal
to the essence of the Talmudic typology, preserving the notion of judicial error as a
departure from an authorized norm. Maimonides’ definitions of the circumstances
applicable to different sorts of error do not correlate precisely to the scholastic approach to tolerable error, but he maintains the distinction between deviation from
a fixed authorized norm and deviation from a norm that is fixed on the base of
secondary principles. In this respect, Maimonides can fairly be characterized as a
fundamentalist, and his approach as consistent with the Talmudic approach, but in
the words immediately following the passage cited above he is revealed as reductionist when he inserts legal reasoning into the category of ta’ut beshiqqul hada’at,
and, even more, when he historicizes the Talmudic typology:37
(And) this was (relevant) before the editing of the Talmud, but in our times the possibility
of this occurring has diminished, for if one issues a ruling, and we find the opposing view in
the Talmud, then he errs in devar mishnah; and if we do not find the opposing view, and his
Certainly, he held the ‘documentation-as-legitimization’ approach. See our discussion on p. 30
below.
34 The fundamentalist character of Maimonides was recently emphasized by Stroumsa (2009,
pp. 53–85).
35 Maimonides, Commentary on the Mishnah. Bekh. 4:4.
36 Maimonides’ reference to the juxtaposition of revelational/derivative law with the terms of nass
and qiyas illustrates a high incorporation of the usul al-fiqh technical terminology.
37 Maimonides, Commentary on the Mishnah. Bekh. 4:4.
33 http://www.springer.com/978-3-319-06583-0